I’ve never tried placing an advert on Facebook, but I’ve been told that FB offers excellent targetting. Apparently it’s possible to require that your ad only be shown to people who fit very narrow definitions, such as people who went to Imperial College and are under 25, or married women living in Cambridge.
Obviously, no amount of matching values from profiles will tell an advertiser exactly who will buy their stuff, but they will want to play the odds. It’s generally a waste of money to advertise a UK service to someone living in France, or to advertise books to someone who hasn’t filled in the “Favourite Books” field. On the other hand, travel, financial products and cosmetic surgery can be usefully adertised to both sexes, so you wouldn’t want to filter out some of that potential audience.
Now, if advertisers know their desired audience, and have tracked their previous successes and failures, they will have a good idea of which fields in my personal information are important and which are not. They probably have hard evidence to back this up.
The second and subsequent lines of my Facebook profile are:
These three bits of info are the truth and the whole (relevant) truth. Facebook advertisers seem to be very interested in them, judging by the ads that are normally shown to me – I guess about a third are dating-related.
So I was a little surprised, initially, to see an ad earlier for www.urbanconnections.co.uk, a gay dating/event site. Didn’t they see that I am “Interested In” women, and not men? A couple of seconds later I clicked that they probably don’t care what I put in “Interested In” – they presumably figure that there are enough men out there in the closet that it’s worth advertising to the apparently straight men as well as the openly gay. And they quite possibly have the statistics to back this up.
I wonder what else you could find out about society by examining Facebook advertising patterns, maybe using multiple fake accounts. Other people have been datamining it recently as well, although not using any advertising-related data. My hope with advertisers’ choices is that at least some of them would be gathering real evidence of what worked, rather than just guessing.